How to make contour smooth

Please, how can I smooth a contour better than the one I attached. The main data is the .xlsx file

X =linspace(-0.3, 0.3, 51);              
Y = linspace(0, 3, 51);       

Answer 1

[Star Strider]

There are a few options, one being to plot fewer levels, and the second being to downsample the matrix using interp2 and also choose fewer levels, and possibly different interpolation method options as well.

Experiment with those approaches here —

figure
imshow(imread('tau_51_51.png'))

M1 = readmatrix('data_tau_51.xlsx');

X = linspace(-0.3, 0.3, 51);              
Y = linspace(0, 3, 51);                  

figure
[c,h] = contourf(X, Y, M1, 6);                                      % Fewer Levels
colormap(turbo)
colorbar

figure
surfc(X, Y, M1);
colormap(turbo)
colorbar
title('Surface Plot of Matrix')

[X1,Y1] = meshgrid(X,Y);


Xr = linspace(min(X), max(X), 26);
Yr = linspace(min(Y), max(Y), 26);
[X2,Y2] = meshgrid(Xr,Yr);

Mr = interp2(X1,Y1,M1,X2,Y2);                                       % Interpolate (Downsample) Matrix

figure
[c,h] = contourf(X2, Y2, Mr, 6);                                    % Fewer Levels Of Interpolated Matrix
colormap(turbo)
colorbar

EDIT — Corrected typographical errors.

.

[University]

Thank you so much. This really helps

[Star Strider]

My pleasure!

If my Answer helped you solve your problem, please Accept it!

.

Answer 2

[Image Analyst]

Looks like you used contourf on some gray scale image. You can get smoother contours by blurring your image before calling contourf. Use imfilter

kernel = ones(9,9) / 9^2;
blurryImage = imfilter(grayImage, kernel);

Answer 3

[John D'Errico]

As others have pointed out, you don't want to try to make the contours smooth AFTER the fact. Instead you want to smooth the surface in advance, THEN build your contours from the smoothed matrix. This will be a better solution. What method of smoothing you employ is a difficult choice, especially since the surface you have appears to have very sharp curvature in places. That will make it difficult no matter what you choose.

[Bruno Luong]

For edge-preserving, usually filter based on anisotropic diffusion can be used

[Image Analyst]

Yes, but @John D'Errico and I believe he doesn't want an edge preserving filter. It's the edges that are giving rise to a blocky edges/outlines. What you want is a blurring filter so the edges will be smoother. But actually I'm not sure why he wants them to be smoother anyway. @University What's wrong with the original, more accurately placed contour lines? Why smooth them out? I don't think it's necessary unless you just like smoother lines for aesthetic reasons.

Again, seeing the original image would help. Please post it.

[Bruno Luong]

@Image Analyst Anisotropic diffussion filter will smooth the data along the edge and keep the normal derivative significantly unchanged. Edge preserving filter will smooth the contours while keeping the density (high at the edge) unmodified. That is exactly what it needs, I think.